Utilization of coherence information from JERS-1/SAR for forest type discrimination

The authors studied the applicability of coherence information which is obtained from interferometric SAR data pair for the discrimination of forest types. Two test sites, around Mt.Fuji in Japan and Phuket Island in Thailand, were tested with two interferometric JERS-1/SAR data pairs for each site. The coherence was computed by complex correlation coefficient in a small corresponding patch between singlelook SAR complex data pair. The temporal changes of SAR intensity and coherence were investigated among various forest types and the results indicated that the coherence information is much superior than the intensity to discriminate different forest types. Especially in Phuket test site, Mangrove forest was easily discriminated from other tropical forest due to its relatively higher coherence, while that forest type is hardly discriminated by using SAR intensity.